Nutritional Assessment Must be Prioritized for Critically Ill Children in the PICU
Although the indirect lung injury group had significantly lower odds of death, these findings were not observed in patients with direct lung injury or in the cohort as a whole (1). Interestingly, the overweight patients with indirect lung injury required longer durations of mechanical ventilation than normal weight patients. The protective effect observed in the obese group was not demonstrated in the overweight. Overall, the association between obesity and outcome in this population is marginal, and any proposed protective effect of obesity on outcomes based on these data seems premature.
Among children included in a large, multicenter registry of cardiac arrests, obesity was associated with lower odds of survival than normal weight (2). Similarly, analysis of our cohort of 1,622 children admitted to PICUs showed that obese patients had significantly higher odds of acquiring an infection, when controlling for covariates (3). Obese patients were also less likely to be discharged from the hospital. Similar to the findings of Ward et al (1), underweight patients in our cohort had the highest risk of remaining hospitalized (3). High quality studies have associated obesity with poor outcomes among children with critical illness, cancer, and stem cell or organ transplants (4).
The lack of a uniform definition of obesity in children is an ongoing problem. Selections of body mass index (BMI) z score cutoff values by Ward et al (1) were intended to assess obesity risk among healthy adolescents and may not transfer to a population that often includes a wide range of acute and chronic conditions affecting growth. The exclusion of children under 2 years old in this study presents an important bias, limiting the generalizability of the findings. A recent study suggested that BMI z score had a higher positive predictive value for obesity at age 2 years than weight for length z score (5).
Ward et al (1) found a relationship between BMI z score and mortality, which was also associated with illness severity (Pediatric Risk of Mortality III) and immune compromise. Other factors not assessed in this study, for example, nutrient intake or premorbid nutritional status may have contributed to this effect, and must be considered in future studies. The physiologic impact of nutritional status on outcomes is an interesting hypothesis. Ward et al (1) emphasize the value of accurate assessment of nutritional status in the PICU environment. We agree with this message. However, their results should not be perceived as a paradoxical advantage for the vulnerable and increasing cohort of critically ill obese patients.